JP2008225871A - Image retrieval device, image retrieval method, and computer program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To quickly and efficiently acquire desired image data without requiring any complicate processing. <P>SOLUTION: When stored image data are called, a display screen on which one or more regions R1, R2 and R3 associated with the image data to a screen display part 2 and the representative (partial) region images R1a, R2a, and R3a of the image data are displayed in parallel is successively sent as shown by (1) → (2) → (3) at fixed intervals, and switched as the representative (partial) region images of the image data, and displayed on the screen display part 2. Thus, the attribute regions R1, R2 and R3 and representative characteristics regions R1a, R2a, and R3a are stored in association with each other, and successively displayed, so that it is possible to identify image data including the similar representative feature regions, and to easily select desired image data. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention saves input document image data, calls and outputs it as necessary, an image forming apparatus such as a digital multifunction peripheral, a document management system, or an image search apparatus used in an image processing apparatus, an image The present invention relates to a search method and a computer program for executing the image search method.

  In general, in an image search device that inputs and saves image data and calls the saved image data when necessary, particularly in an image search device that searches for similar images, data display means such as an image If the file is not large enough, the user can check the stored data by displaying a list of information such as name, date, registrant, file size, or images such as thumbnails. There is a method for enabling an intuitive selection using the entire reduced image. However, when the reduced image is used, it may be impossible to distinguish between images having similar structures in appearance (see FIG. 7).

  When such a reduced image is used, for example, an invention described in Patent Document 1 or 2 is known as a technique for distinguishing between images having similar structures in appearance. In the invention described in Patent Document 1, a reduced image creating unit reads electronic data such as image data, a document created by a word processor, spreadsheet data, and the like, and creates a reduced image. The similarity between the reduced images is obtained, and if the similarity is larger than a predetermined threshold, the difference image creation unit cuts out the region of the difference portion in the similar images and displays the portion on the image display unit. That's it. In the invention described in Patent Document 2, document layout means for extracting each layout of a plurality of documents and sorting the documents based on the layout, and dissimilar location extraction for extracting dissimilar locations in each document And a dissimilar location display means for displaying only the locations extracted by the dissimilar location extraction means, and a similar location extraction means for extracting similar locations in each document, and a similar location extraction means. It has similar location display means for displaying only the extracted locations. In these inventions, if the reduced images are similar, the partial images obtained by dividing the original image are extracted and displayed with low similarity. According to this, even if the reduced images of the entire image data are similar and difficult to discriminate, it is possible to distinguish because the portion with a large difference is displayed.

In addition, there is an invention described in Patent Document 3 as a method for distinguishing between images having similar reduced images. The present invention provides a document file acquisition unit that acquires a document file, a reduced image generation unit that generates a reduced image of an image related to the document file acquired by the document file acquisition unit, and the reduced image displayed on the display is designated by a pointer. A processing area determining means for determining a processing area on the image related to the document file, a processing area information extracting means for extracting information in the processing area from the document file, and a processing area information extracting means. The processing area information display means for displaying the extracted processing area information on the display in a predetermined size. When the operator designates a focused part of the displayed reduced image, the area around the designated part is displayed. Extract the area from the image data that is the basis of the reduced image, extract information from the area, and output and display the extracted information It is. According to this, it is possible to distinguish between image data having similar reduced images by displaying information of a portion of interest.
JP 2003-331299 A JP 2005-063220 A JP 2005-017592 A

  However, in the inventions described in Patent Documents 1 and 2, when a large number of search results are to be displayed, a reduced image of the search results is displayed because the size of the display device is limited. Difficult to grasp. Therefore, the techniques described in Patent Documents 1 and 2 are known in order to display a feature region image as a search result.

  However, the feature region image in the techniques described in Patent Documents 1 and 2 is an image having the smallest similarity (not similar) to a partial image in another image displayed as a search result. There are cases in which the image is merely a relative feature image in the relationship and is not a characteristic region of the original image. Furthermore, if there are many images displayed as search results, the amount of calculation increases accordingly, which causes a reduction in processing speed.

  In addition, in the invention described in Patent Document 3, in order to efficiently specify which part of a similar reduced image should be distinguished, it is possible to specify in advance, which information is likely to be different, empirical prediction, etc. If you do not know it, you will specify the part you want to pay attention to, and it is inefficient. Furthermore, after designating the part of interest, the original image data area is extracted and processed, so that the processing time is waited. If the processing contents become complicated, the waiting time will increase, which is inconvenient.

  Therefore, a problem to be solved by the present invention is to obtain desired image data efficiently in a short time without requiring complicated processing.

  In order to solve the above-mentioned problem, the first means is searched by the means for searching for an image similar to the image data and the means for searching in an image search device for searching for an image similar to the specified image data. Means for extracting a feature region image stored in association with the image using the image as a key, and displaying the feature region image on a display screen, wherein the means for displaying has a predetermined order in accordance with the type of the feature region image. A plurality of feature region images are displayed on the display screen at time intervals.

  A second means is the image search apparatus according to the first means, wherein the display means displays the image in a size corresponding to the type of the feature area image.

  According to a third means, in the first or second means, the means for displaying distinguishes between the feature area image to be displayed and the specific area image not to be displayed according to the type of the feature area image, and only the feature image to be displayed is displayed. Is displayed.

  According to a fourth means, in any one of the first to third means, when the means for displaying obtains position information of a representative feature area, and displays a plurality of representative feature areas by switching. The feature region order to be displayed is determined by position information.

  The fifth means is one of the first to third means, in which the means for displaying obtains the size information of the representative feature area and displays a plurality of representative feature areas by switching. The order of feature regions to be displayed is determined by size information.

  Sixth means, in any one of the first to fifth means, when the means for displaying obtains attribute information of a representative feature region, and switches and displays a plurality of representative feature regions, The feature region order to be displayed is determined by attribute information.

  The seventh means is characterized in that in any one of the first to sixth means, the representative feature area is scaled and displayed to a size that can be displayed on the display screen.

  In an image search method for searching for an image similar to the designated image data, an eighth means is a step of searching for an image similar to the image data, and is stored in association with the image using the searched image as a key. And a step of displaying a plurality of the feature area images on a display screen at predetermined time intervals in an order corresponding to the type of the extracted feature area image. An image search is performed based on the feature region image.

  The ninth means is characterized in that when the feature area image is displayed in the step of displaying the feature area image in the eighth means, the image is displayed in a size corresponding to the type of the feature area image. retrieval method.

  The tenth means, when displaying the feature area image in the step of displaying the feature area image in the eighth or ninth means, does not display the feature area image to be displayed according to the type of the feature area image. It is characterized in that only the characteristic image to be displayed is displayed while being distinguished from the specific area image.

  The eleventh means obtains positional information of representative feature areas when displaying the feature image area in the step of displaying the feature area image in any of the eighth to tenth means, When the representative feature areas are switched and displayed, the order of the feature areas to be displayed is determined by the position information.

  The twelfth means obtains representative feature area size information when displaying the feature image area in the step of displaying the feature area image in any of the eighth to tenth means, When a plurality of representative feature areas are switched and displayed, the order of the feature areas to be displayed is determined by size information.

  The thirteenth means obtains attribute information of a representative feature area when displaying the feature image area in the step of displaying the feature area image in any of the eighth to twelfth means. When the representative feature areas are switched and displayed, the order of the feature areas to be displayed is determined by the attribute information.

  In a fourteenth means, in any one of the eighth to thirteenth means, in the step of displaying the feature area image, the representative feature area is scaled to be displayed on the display screen and displayed. Features.

  The fifteenth means is characterized by a computer program comprising a procedure for executing the steps in the image search method according to any one of the eighth to fourteenth means by a computer.

  In the embodiment described later, the means for searching for and displaying the image similar to the image data corresponds to the central processing unit 8, the display screen corresponds to the image display unit 2, and the feature region image is searched for by the means for searching. The retrieved image is stored in the auxiliary storage device 6 in association with the image as a key, and the feature region image is extracted by the feature region extraction unit 4.

  According to the present invention, a plurality of representative feature regions are extracted from input image data, and a plurality of feature region images are displayed on the display screen at predetermined time intervals in an order corresponding to the type of the feature region image. Therefore, it is possible to obtain desired image data efficiently in a short time without requiring complicated processing.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  FIG. 1 is a block diagram showing a schematic configuration of an image search apparatus according to an embodiment of the present invention. In the figure, an image search apparatus is for image data such as an image input unit 1 such as a scanner, an input image data such as a liquid crystal display or an image display unit 2 that displays stored image data, a mouse or a touch panel. An information input unit 3 for inputting information such as a designated region and a point of interest, a feature region extracting unit 4 for extracting a representative feature region from image data, a main memory 5 for temporarily storing data to be processed, and associating with image data It comprises an auxiliary storage device 6 that stores a plurality of representative feature areas in association with each other, a communication device 7 that is used when data is exchanged with the outside through a network, and a central processing unit 8 that controls processing.

  2 is a flowchart showing the processing procedure of the image processing apparatus executed by the central processing unit 8 of FIG. 1. FIG. 2 (a) shows the procedure for inputting image data, and FIG. 2 (b) shows the procedure for outputting image data. Show.

  First, in FIG. 2A, when an image is input (step S101), the feature area extraction unit 4 analyzes the input image data by analyzing the layout (step S102) and cutting out the area (step S103). ), Attribute identification (step S104) for identifying the cut out region as a sentence, title, figure, photograph, table, or the like is sequentially executed to extract a representative feature region. The layout analysis, area segmentation, and area attribute identification are not particularly described here. For example, a known method disclosed in JP-A-7-37036 or JP-A-9-114923 can be used. Can be used.

  In general, when selecting desired image data from a plurality of image data by visual inspection, it is easier to understand the contents of the entire document than by understanding the contents of the document, photos, titles, etc. There are many cases. Therefore, as representative feature areas, as shown in FIG. 6, attributes (types) such as FIG. R1, a photograph, a table R2, and a title R3 are selected from the areas cut out from the original 10, and the representative feature areas. Are stored in association with the image data. In addition to the auxiliary storage device 6, the image data can be stored in another auxiliary storage device connected to the network via a communication device. R4 represents a character area.

  FIG. 7 is a diagram showing a state in which the original 10 shown in FIG. 6 and the entire original 10 ′ similar to the original 10 are reduced and arranged side by side on the image display unit (display screen) 2. Since the layout is the same from this display, it is difficult to understand the contents of this manuscript (document) without looking at the details of the manuscript. Therefore, in the present embodiment, the display is as follows.

  That is, when the stored image data is called up, one or more regions R1, R2, R3 associated with the image data with respect to the image display unit 2 and a representative (partial) region image of the image data A display screen on which R1a, R2a, and R3a are displayed in parallel is sequentially sent in the order of (1) → (2) → (3) at regular intervals, and as shown in FIG. The image is displayed on the image display unit 2 by switching as a typical (partial) region image. As described above, the attribute regions R1, R2, R3 and the representative feature regions R1a, R2a, R3a are stored in association with each other, and are sequentially displayed, so that image data including similar representative feature regions can be displayed. Identification can be made possible and selection of desired image data can be facilitated. FIG. 2B shows the processing procedure at this time. The image data is loaded (step S201), the display of the feature region is updated (step S202), and sequentially displayed (step S203). Is selected (step S204) and output (step S205).

  As shown in FIG. 8, a plurality of attribute areas R1, R2, and R3 and representative feature areas R1a, R2a, and R3a can be switched and displayed in accordance with an instruction from an operator instead of being switched and displayed at a fixed time. The processing procedure at this time is shown in FIG. FIG. 3 is a flowchart showing a processing procedure for switching and displaying according to an instruction from the operator.

  In this processing procedure, after the feature area is displayed in step S203 of the processing procedure shown in FIG. 2B, when an instruction is given by the operator, the display is switched (step S301), and the process returns to step S202. It is set so that the display of the feature amount is again updated and sequentially displayed.

  When processing is performed in this way, when there are a large number of representative feature regions, there is a case where switching in a certain time cannot be waited. Desired image data can be selected more efficiently by the operator instructing the switching timing.

  An image is input from step S101. Depending on the content of the input image data, there is an error in a representative feature region as a result of layout analysis, region extraction, or region attribute identification, or an unnecessary figure is displayed. May include. For this reason, after extracting representative feature regions, they are sequentially displayed on the display means, and the operator is prompted to select whether or not to store them in association with image data. After the operator finishes selecting the extracted regions as all representative feature regions, the image data and the selected representative feature regions are stored in association with each other.

  The processing procedure at this time is shown in FIG. FIG. 4 is a flowchart showing a processing procedure of processing for selecting whether or not an operator should store a representative feature region in association with image data. In FIG. 4, when an image is input (step S <b> 101), layout analysis (step S <b> 102), region extraction (step S <b> 103), and extraction are performed on the input image data by the feature region extraction unit 4. Attribute identification (step S104) for identifying the area as text, title, figure, photo, table, or the like is sequentially executed to extract a representative feature area. Next, after extracting a representative feature region, it is sequentially displayed on the display means (step S401). The operator selects whether or not to store the image data in association with the image data (step S402), and when the selection of the regions extracted as all the representative feature regions is completed (step S403), the image data and the representative selected The feature area is stored in association with each other (step S105 ′).

  As described above, by discarding representative feature areas, it is possible to reduce errors and unnecessary areas, and it is possible to efficiently select desired image data.

  In addition, as described above, an image is input from step S101. Depending on the content of the input image data, an extraction leak may be detected in a representative feature region as a result of layout analysis, region extraction, and region attribute identification. May include. Therefore, after extracting a representative feature area, it is sequentially displayed on the display means, and the operator is prompted to determine whether or not the representative feature area is designated by the information input means. If the area is designated, the image data is displayed on the display means, the area to be designated is inputted by the information input means, and the area obtained thereby is saved as a representative feature area in association with the image data.

  The processing procedure at this time is shown in FIG. FIG. 5 is a flowchart showing a processing procedure of processing for selecting whether or not the operator designates with the information input means. In FIG. 4, when an image is input (step S <b> 101), layout analysis (step S <b> 102), region extraction (step S <b> 103), and extraction are performed on the input image data by the feature region extraction unit 4. Attribute identification (step S104) for identifying the area as text, title, figure, photo, table, or the like is sequentially executed to extract a representative feature area. Next, after extracting representative feature regions, they are sequentially displayed on the display means (step S501). The operator selects whether or not to add a representative feature area (step S502). When all the additions are completed (step S503 → S502), the image data and the selected representative feature area are stored in association with each other. (Step S105 ′).

  In this way, by allowing the operator to designate a representative feature region, it is possible to prevent a representative feature region from being extracted, and it is possible to efficiently select desired image data.

  By the way, in each process as described above, a representative feature region is obtained as a result of layout analysis in step S102, region extraction in step S103, and region attribute identification in step S104. Therefore, data can be identified by performing various sorts according to the characteristics of these characteristic regions. For example, the position, size, and type of a specific portion such as the center of gravity of the feature region.

1) When sorting in the order of the position of a specific part such as the center of gravity of the feature area, and the display order when switching and displaying the representative feature area is to follow the sorted order
In this case, sorting is performed in the order of the positions of specific parts such as the center of gravity of the feature area, and the display order when the representative feature areas are switched and displayed is made to follow the sorted order. In general, in a horizontally written document image, the line of sight moves from the upper left to the right, and then gradually looks downward, but if representative feature areas are displayed in that order, they are represented in the order in which they appear when reading the document image data. A characteristic region can be displayed, which may distinguish image data.

  Displaying representative feature areas in the order according to the position information in this way helps to intuitively understand the contents of the image data, and allows the desired image data to be selected efficiently.

2) When sorting in the order of the size of the feature area and switching the display of representative feature areas to follow the sorted order In this case, the area of the feature area, etc. Sorting is performed in order of size, and the display order when switching and displaying representative feature areas is made to follow the sorted order. In general, a larger representative feature area included in image data is more conspicuous. Therefore, image data may be distinguished by a feature region that is larger than a small feature region among representative feature regions included in the image data.

  Displaying representative feature areas in the order corresponding to the size information in this way helps to intuitively understand the contents of the image data, and allows the desired image data to be selected efficiently.

3) Sorting in order of attributes such as the type of feature area, etc., so that the display order when switching and displaying the representative feature area follows the sorted order In this case, the type of feature area The attributes are sorted in the order of the attributes, and the display order when the representative feature areas are switched and displayed is made to follow the sorted order. In general, image data may be distinguished by a more impressive attribute such as a photograph or a drawing than character information such as a title among representative feature areas included in the image data.

  Displaying the representative feature areas in the order according to the attribute information in this way helps to intuitively understand the contents of the image data, and allows the desired image data to be selected efficiently.

  In FIG. 2, in step S202, a representative feature area is sent (updated) to the display means at regular intervals and displayed, but the display time can be changed according to the attribute of the feature area. If the feature area includes characters such as a title as compared to a figure or photo, the content is read, and the content may not be confirmed in the same display time as the figure or photo. Therefore, the attribute identified at the time of input in step S104 is checked, and if the attribute is an area including characters, the display time is set longer. This secures time for reading characters and helps to understand the contents of image data. As a result, desired image data can be selected efficiently. It should be noted that the display order or time of the feature area images can be arbitrarily changed by the user.

  By the way, in the display of the typical feature region of the image display unit 2, the display size is displayed as it is as shown in FIG. 8, but the display size can be changed. In general, many image processing (forming) devices such as MFPs have only a small display device (image display unit 2). Therefore, there is a possibility that the feature area which is a partial area of the image data does not fit in the displayable area of the display device. On the other hand, there is a possibility that the characteristic area is small compared to the displayable size of the display device and is difficult to discriminate. Therefore, as shown in FIG. 9, the attribute area and the representative feature area are scaled to a size that can be displayed on the image display unit 2 and displayed in an easy-to-view manner. Thereby, desired image data can be selected efficiently. In the example shown in FIG. 9, the enlarged attribute areas are indicated by R1b, R2b, R3b, and the representative feature areas are indicated by R3c, R1c, R2c. Also in this case, the display is switched in the order of (1) → (2) → (3).

  As shown in FIG. 9, when the representative feature areas R1c, R2c, and R3c are scaled to a size that can be displayed by the image display unit 2, the relative sizes of the areas R1, R2, and R3 are displayed. It is also possible to change the magnification as shown in FIG. 10 while preserving the ratio. In other words, each attribute region and feature region are scaled at a magnification that scales the largest representative feature region R3 into an image R3b of a size that can be displayed on the image display unit 2 (a frame 2a indicated by a chain line). In the figure, a figure region R1d and a table region R2d obtained by scaling the title region R3 with a scaling factor, and representative feature regions R1e and R2e that are scaled with the same magnification corresponding to the region are shown. As a result, the difference in the relative size of the original image is maintained even after scaling, so that the image data may be intuitively distinguished by the difference in size between the representative feature areas. However, a lower limit of magnification is necessary to prevent the display size from becoming too small and making it impossible to discriminate the image. In this case as well, the display is switched in the order of (1) → (2) → (3).

  By storing the ratio of the size of the representative feature regions in this way, it is possible to more intuitively understand the contents of the image data and to efficiently select desired image data.

  When a representative feature area is displayed in a size that can be displayed by the display means, the scaling factor is changed according to the attribute of each representative feature area. In particular, in the case of a representative area including characters such as a title, it tends to be difficult to discriminate when the display size is smaller than in a figure or photo area. Therefore, in the case of an area having an attribute that includes characters, the magnification is changed to be larger than other areas. When the attribute is an area including characters, increasing the display magnification helps to understand the contents of the image data, and allows desired image data to be selected efficiently.

  Also, it is possible to display a list of all typical feature areas of designated image data on the image display unit 2 in accordance with the operation of the operator. By adding such a mode for displaying a list, it is possible to intuitively understand the contents of the image data and to select desired image data efficiently.

  Note that a reduced image of the entire image data as shown in FIG. 7 can be added as one of the representative feature areas to be displayed. By adding a reduced image, not only the representative feature areas are sequentially switched and displayed, but also information regarding their arrangement can be obtained. This helps to understand the contents of the image data and allows the desired image data to be selected efficiently.

As described above, according to this embodiment,
(1) Display means between image data including similar representative feature areas by extracting and holding a plurality of representative feature areas included in the image data and sequentially switching and displaying them Can be identified, and selection of desired image data can be facilitated.
(2) Desired image data can be selected more efficiently by sequentially switching and displaying a plurality of representative feature regions in accordance with an instruction from the operator.
(3) By selecting the representative feature areas extracted by the feature area extraction means, it becomes possible to reduce errors and unnecessary areas, and it is possible to efficiently select desired image data.
(4) By allowing a representative feature region to be designated by an operator, it is possible to prevent a typical feature region from being extracted, and it is possible to efficiently select desired image data.
(5) Representative feature areas are displayed in an order according to position information, representative feature areas are displayed in an order according to size information, or representative features are displayed in an order according to attribute information By displaying the area, it becomes possible to help the contents of the image data more intuitively and desired image data can be selected efficiently.
(6) When the attribute is an area including characters, it helps to understand the contents of the image data by increasing the display time or by increasing the display magnification. Desired image data can be selected.
(7) It is possible to efficiently select desired image data by scaling the representative feature area to a size that can be displayed by the display means and displaying it easily.
(8) By displaying the ratio of the size of the representative feature area while saving, it helps to understand the contents of the image data more intuitively, and the desired image data can be selected efficiently. .
(9) By adding a mode for displaying a list, it is possible to help intuitively understand the contents of image data, and to select desired image data efficiently.
(10) By adding a reduced image of the entire image data as one of the representative feature areas, it is possible to help understand the contents of the image data and to select desired image data efficiently.
There are effects such as.

1 is a block diagram illustrating a schematic configuration of an image search device according to an embodiment of the present invention. It is a flowchart which shows the process sequence of the said image processing apparatus performed by the central processing unit of FIG. It is a flowchart which shows the process sequence switched and displayed by the instruction | indication from an operator. It is a flowchart which shows the process sequence of the process which the operator selects whether it should preserve | save a typical characteristic area in association with image data. It is a flowchart which shows the process sequence of the process which selects whether the operator designates with an information input means. It is a figure which shows the relationship between the original and the typical characteristic area cut out from the original. It is a figure which shows the state which displayed the reduced image of the whole similar image data side by side. It is a figure which shows the state which arranged the typical characteristic area | region of the similar image data, and switched and displayed. FIG. 9 is a diagram illustrating a state in which the image data displayed in FIG. FIG. 9 is a diagram illustrating a state in which the image data displayed in FIG. 8 is scaled while maintaining a relative size ratio.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Image input part 2 Image display part 3 Information input part 4 Feature area extraction part 5 Main memory 6 Auxiliary storage device 7 Communication apparatus 8 Central processing unit

Claims (15)

In an image search device for searching for an image similar to specified image data,
Means for searching for an image similar to the image data;
Means for extracting a feature region image stored in association with the image using the image searched by the searching means as a key, and displaying the feature region image on a display screen;
With
The image search apparatus characterized in that the display means displays the plurality of feature region images on the display screen at a predetermined time interval in an order corresponding to the type of the feature region image. The image search device according to claim 1,
The image search apparatus according to claim 1, wherein the display means displays a size corresponding to a type of the feature area image. The image search device according to claim 1 or 2,
The image search apparatus characterized in that the means for displaying distinguishes between the feature area image to be displayed and the specific area image not to be displayed according to the type of the feature area image, and displays only the feature image to be displayed. The image search device according to any one of claims 1 to 3,
The display means obtains position information of representative feature areas, and determines the order of the feature areas to be displayed when switching and displaying a plurality of representative feature areas based on the position information. Image search device. The image search device according to any one of claims 1 to 3,
The means for displaying obtains the size information of a representative feature region, and when the plurality of representative feature regions are switched and displayed, the order of the feature regions to be displayed is determined by the size information. An image search device. The image search device according to any one of claims 1 to 5,
The means for displaying obtains attribute information of a representative feature area, and when displaying a plurality of representative feature areas by switching, the order of the feature areas to be displayed is determined by the attribute information. Image search device. The image search device according to any one of claims 1 to 6,
An image search apparatus characterized in that a representative feature area is displayed in a size that can be displayed on the display screen. In an image search method for searching for an image similar to specified image data,
Searching for an image similar to the image data;
Extracting a feature region image stored in association with the image using the searched image as a key;
Displaying the plurality of feature region images on a display screen at predetermined time intervals in an order according to the type of the extracted feature region image;
An image search method comprising: performing an image search based on the displayed feature region image. The image search method according to claim 8.
An image search method, wherein when displaying the feature region image in the step of displaying the feature region image, the feature region image is displayed in a size corresponding to a type of the feature region image. The image search method according to claim 8 or 9,
When displaying the feature region image in the step of displaying the feature region image, according to the type of the feature region image, the feature region image to be displayed is distinguished from the specific region image not to be displayed, and only the feature image to be displayed is displayed. An image search method characterized by: The image search method according to any one of claims 8 to 10,
When displaying the feature image region in the step of displaying the feature region image, the position information of the representative feature region is obtained, and when displaying a plurality of representative feature regions, the feature region to be displayed is displayed. An image search method characterized in that a rank is determined by position information. The image search device according to any one of claims 8 to 10,
When displaying the feature image area in the step of displaying the feature area image, the size information of the representative feature area is obtained, and the feature area to be displayed is displayed when switching the plurality of representative feature areas. The image search method is characterized in that the ranking is determined by size information. The image search method according to any one of claims 8 to 12,
When displaying the feature image area in the step of displaying the feature area image, the attribute information of the representative feature area is obtained, and when displaying a plurality of representative feature areas, the feature area to be displayed is displayed. An image search method characterized in that a rank is determined by attribute information. The image search device according to any one of claims 8 to 13,
An image search method characterized in that in the step of displaying the feature region image, the representative feature region is scaled and displayed to a size that can be displayed on the display screen.   15. A computer program comprising a procedure for executing the steps in the image search method according to claim 8 by a computer.

Images